Repository navigation
Expand file tree
/
Copy pathllm.py
More file actions
218 lines (186 loc) · 7.45 KB
/
Copy pathllm.py
File metadata and controls
218 lines (186 loc) · 7.45 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
#!/usr/bin/env python3
"""
DevFlow - LLM 桥接层
统一封装 LLM 调用,供 explain / testgen / agents 等脚本的可选 --llm 模式使用。
复用 model-manager.py 的配置(~/.devflow/models.json + 环境变量)。
设计原则:
- 零第三方依赖:使用标准库 urllib 调用 OpenAI 兼容 /chat/completions 端点
- 若已安装 openai/anthropic SDK 则优先使用(可选增强)
- 未配置 API key 时 is_available() 返回 False,调用方自动回退到启发式
Usage:
from llm import is_available, chat, get_active_provider
"""
import json
import os
import urllib.request
from pathlib import Path
CONFIG_DIR = Path.home() / ".devflow"
CONFIG_FILE = CONFIG_DIR / "models.json"
PROVIDERS = {
"openai": {
"name": "OpenAI",
"env_key": "OPENAI_API_KEY",
"base_url": "https://api.openai.com/v1",
"model": "gpt-4o-mini",
},
"claude": {
"name": "Claude (Anthropic)",
"env_key": "ANTHROPIC_API_KEY",
"base_url": "https://api.anthropic.com/v1",
"model": "claude-3-haiku-20240307",
},
"local": {
"name": "Local Model (Ollama)",
"env_key": None,
"base_url": "http://localhost:11434",
"model": "llama3",
},
"azure": {
"name": "Azure OpenAI",
"env_key": "AZURE_API_KEY",
"base_url": "",
"model": "gpt-4o-mini",
},
}
def _load_config():
if CONFIG_FILE.exists():
try:
return json.loads(CONFIG_FILE.read_text(encoding="utf-8"))
except Exception:
pass
return {"default_provider": "openai", "providers": {}}
def get_active_provider():
"""返回当前激活的 provider id"""
cfg = _load_config()
return cfg.get("default_provider", "openai")
def get_api_key(provider=None):
"""获取 API key:优先环境变量,其次配置文件"""
provider = provider or get_active_provider()
env_key = PROVIDERS.get(provider, {}).get("env_key")
if env_key:
key = os.environ.get(env_key)
if key:
return key
cfg = _load_config()
return cfg.get("providers", {}).get(provider, {}).get("api_key")
def get_model(provider=None):
"""获取模型名:优先配置文件,其次默认"""
provider = provider or get_active_provider()
cfg = _load_config()
configured = cfg.get("providers", {}).get(provider, {}).get("default_model")
if configured:
return configured
return PROVIDERS.get(provider, {}).get("model")
def get_base_url(provider=None):
"""获取 base_url:优先配置文件,其次默认"""
provider = provider or get_active_provider()
cfg = _load_config()
configured = cfg.get("providers", {}).get(provider, {}).get("base_url")
if configured:
return configured.rstrip("/")
return PROVIDERS.get(provider, {}).get("base_url", "").rstrip("/")
def is_available(provider=None):
"""LLM 是否可用:local 检查端口,其余检查 API key"""
provider = provider or get_active_provider()
if provider == "local":
return True # 假设本地服务可能运行;实际调用失败会抛异常
return bool(get_api_key(provider))
def _chat_openai_compatible(base_url, api_key, model, messages, max_tokens):
"""调用 OpenAI 兼容 /chat/completions 端点"""
url = base_url.rstrip("/") + "/chat/completions"
payload = json.dumps({
"model": model,
"messages": messages,
"max_tokens": max_tokens,
"temperature": 0.2,
}).encode("utf-8")
req = urllib.request.Request(url, data=payload, method="POST")
req.add_header("Content-Type", "application/json")
if api_key:
req.add_header("Authorization", f"Bearer {api_key}")
with urllib.request.urlopen(req, timeout=60) as resp:
data = json.loads(resp.read().decode("utf-8"))
return data["choices"][0]["message"]["content"]
def _chat_ollama(base_url, model, messages, max_tokens):
"""调用本地 Ollama /api/chat 端点"""
url = base_url.rstrip("/") + "/api/chat"
payload = json.dumps({
"model": model,
"messages": messages,
"stream": False,
"options": {"num_predict": max_tokens},
}).encode("utf-8")
req = urllib.request.Request(url, data=payload, method="POST")
req.add_header("Content-Type", "application/json")
with urllib.request.urlopen(req, timeout=120) as resp:
data = json.loads(resp.read().decode("utf-8"))
return data["message"]["content"]
def _chat_anthropic(base_url, api_key, model, messages, max_tokens):
"""调用 Anthropic Messages API"""
url = base_url.rstrip("/") + "/messages"
# 转换 OpenAI 消息格式 → Anthropic 格式
system_msgs = [m["content"] for m in messages if m["role"] == "system"]
convo = [m for m in messages if m["role"] != "system"]
payload = json.dumps({
"model": model,
"max_tokens": max_tokens,
"system": "\n".join(system_msgs) if system_msgs else None,
"messages": convo,
}).encode("utf-8")
req = urllib.request.Request(url, data=payload, method="POST")
req.add_header("Content-Type", "application/json")
req.add_header("x-api-key", api_key)
req.add_header("anthropic-version", "2023-06-01")
with urllib.request.urlopen(req, timeout=60) as resp:
data = json.loads(resp.read().decode("utf-8"))
return "".join(block.get("text", "") for block in data.get("content", []))
def chat(prompt, system=None, provider=None, max_tokens=2048):
"""发送对话,返回 LLM 文本响应。
参数:
prompt: 用户消息
system: 可选的 system 提示
provider: 显式指定 provider(默认用配置的)
max_tokens: 最大输出 token
返回:
str 响应文本
异常:
RuntimeError / urllib 错误
"""
provider = provider or get_active_provider()
messages = []
if system:
messages.append({"role": "system", "content": system})
messages.append({"role": "user", "content": prompt})
api_key = get_api_key(provider)
model = get_model(provider)
base_url = get_base_url(provider)
if provider == "local":
return _chat_ollama(base_url, model, messages, max_tokens)
if provider == "claude":
if not api_key:
raise RuntimeError(f"未配置 {PROVIDERS['claude']['name']} API key")
return _chat_anthropic(base_url, api_key, model, messages, max_tokens)
# openai / azure / 其他 OpenAI 兼容端点
if not api_key:
raise RuntimeError(f"未配置 {PROVIDERS.get(provider, {}).get('name', provider)} API key")
return _chat_openai_compatible(base_url, api_key, model, messages, max_tokens)
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser(description="DevFlow LLM 桥接层")
parser.add_argument("--check", action="store_true", help="检查 LLM 是否可用")
parser.add_argument("--provider", help="指定 provider")
parser.add_argument("--prompt", help="发送一条提示")
args = parser.parse_args()
if args.check:
prov = args.provider or get_active_provider()
print(f"provider: {prov}")
print(f"model: {get_model(prov)}")
print(f"base_url: {get_base_url(prov)}")
print(f"available: {'✅' if is_available(prov) else '❌'}")
elif args.prompt:
try:
print(chat(args.prompt, provider=args.provider))
except Exception as e:
print(f"LLM 调用失败: {e}", file=__import__("sys").stderr)
else:
parser.print_help()